For reasons that ultimately led to my cryocooler purchase, I wanted to have a source of quite high pressure (~1000 psi) air. Most ordinary utility air compressors top out at about 150 psi, which, working in the oil and gas industry, seems like small potatoes. I wanted some pressure, I'm talking ANSI 600 class at least. In my line of work, we use devices called blowcases to pump liquids using high pressure gas. They're convenient when you need to pump a liquid to relatively high pressures and don't have any utility electric power, but do have a source of high pressure gas. I figured I could do this in reverse, taking a source of high pressure liquid to pressurize a gas.
The "compressor" works by allowing gas from a lower pressure source (a shop air compressor capable of 150 psi discharge pressure) to enter the compression chamber. Water is then pumped in from the bottom, compressing the gas and allowing it to flow into a reservoir. Once the water reaches a high-level sensing line, the flow of water in is stopped and the water is drained out the bottom. The cycle is then repeated until the pressure in the reservoir is the desired pressure. The level is sensed by a magnet wire inserted into the chamber through a hole drilled in a plug, and then epoxied in place. An Arduino was used to sense when the resistance between the chamber walls and the sensing wire fell, indicating that the water level had reached the wire (good thing mildly impure water is a good conductor). The Arduino then switches a pair of automated valves via a solenoid valve to shut off flow from the pressure washer and drain the chamber. After a set time, the valves are switched again and the cycle repeats. Here's a terribly shaky video explanation of the contraption. Yes, that (was) my kitchen. No, that floor no longer haunts my dreams. Never buy a house, but that's another anthology.
A nice thing about living in an oil field (there are a small handful) is that the local plumbing supply store sells Schedule 80 A106 Grade B pipe capable of withstanding well over 1000 psi and 3000 psi rated fittings. I bought some 2" pipe nipples, and couplings, bushings, and tees. From McMaster Carr (truly the greatest retailer in the solar system), I bought some 1/4" NPT fittings and steel tubing. Off of ebay, I found some reasonably priced 1000 and 2000 psi rated ball valves equipped with pneumatic actuators and an Asco solenoid valve. One happened to be installed as fail open and the other as fail closed, which made it possible to control both valves with the single solenoid valve (although it's pretty easy to switch the fail position on pneumatic actuators of this type). Hooking the pressure washer up to the 2000 psi actuated valve and using the 1000 psi valve to drain the chamber, the contraption actually worked to get up to beyond 900 psi!
|pics or it didn't happen|
Granted, it did turn out to be quite a pain in the butt to set up, and the time to pressurize the air was quite long, so why in the world of alabaster did I want to do this? At the time, I was thinking the easiest way I could generate liquid air was to allow high pressure air to go through a Joule Thompson expansion to cool it down to liquefying temperatures. However, to JT straight to a liquid, you need to get on top of the liquid envelope on the pressure-enthalpy diagram, that is to say pressurize it beyond its critical pressure, which for air is about 37 atmospheres or 550 psi. After that, you still need to chill it down to about 150K and then JT it to get it into the liquid envelope, but I figured 150K would be easier to achieve than 80K. Turns out, not really, I never got a way to get to those temperatures until I got my cryocooler, which could get all the way to 80 K anyways. It was a fun proof of concept nonetheless.